Abstarct: The improvement of seismic data resolution is very important in hydrocarbon exploration, especially in small scale structures like thin laxyers that they are places of passing and accumulation of hydrocarbon. Because this structures are below of the seismic vertical resolution (tuning thickness), so to detect of this type reservoirs, high resolution seismic data are needed. Many methods have been introduced to increase the resolution of seismic data. Different deconvolution methods are common methods of vertical resolution enhancement. In this methods baxsed on the convolution model, the seismic trace is a convolution of the reflectivity series with a seismic wavelet. The seismic wavelet has often frequency band-limited, whilst the frequency bandwidth of reflectivity series is unlimited. Through the convolution process, seismic wavelet is compressed in time domain and improves the seismic resolution. Since in this methods, is considered assumptions baxsed on convolution model of reflection seismic data which have often little reality, therefore, many researchers tried to increase the vertical resolution data using other methods. Including new methods of increasing the vertical resolution can be mentioned to methods baxsed on continuous and discrete wavelet transform. In this thesis, increase the vertical resolution of seismic data using time scaling property of Fourier transform, This means that a filter is designed baxsed on Fourier transform of scaled and original source wavelet that by applying on seismic data, increases the frequency band and improves resolution. The algorithm is tested on synthetic and real seismic data and the results are compared with frequency deconvolution method. The results showed that the method presented’s performance better than the conventional deconvolution in frequency domain.Since seismic reflection data has got time-varying frequency content, therefore in the following is designed the scale filter in time-frequency domain to improve the method, in order to enhance the resolution of nonstationary seismic data. In this case, the considered seismic data using a time-frequency transform as Gabor transform is extended to 2-D time-frequency domain and the effect of scale transform to increase the content of frequency and thus improving the resolution of seismic is used. The gained result of the applied method on synthetic and real data has proved that the proposed method improves resolution by expanding frequency bandwidth. The results the new method in compare with Gabor deconvolution is shown to be more efficient.